On Wed, 2022-12-21 at 14:58 +0100, Thibaut Lunet wrote: > Hi everyone, > > I want to vectorize multiple matrix-vector products and avoid a for > loop, a little bit like np.linalg.solve does, for instance : >
<snip> > > dimension > 2. > > Since np.linalg.solve does this vectorization naturally, I wonder if > there is a way to get the same behavior with matrix-vector > multiplication already in numpy, and why np.matmul does not behave > like > np.linalg.solve does ? All of these functions need some convention to deal with ambiguity of matrix vs. stack of vector. Since we don't indicate it on the array itself, both matmul and the `@` operator assume matrix inputs unless given a vector. There have been thoughts of having a pair of `vecmat` and `matvec` functions to make working with stacked vectors more convenient. I don't think anyone ever pushed for it very strongly. There is maybe a bit more of a push for `vecdot` (both vectors) right now, but you want the mixed case... - Sebastian > > Best, > > Thibaut > > _______________________________________________ > NumPy-Discussion mailing list -- numpy-discussion@python.org > To unsubscribe send an email to numpy-discussion-le...@python.org > https://mail.python.org/mailman3/lists/numpy-discussion.python.org/ > Member address: sebast...@sipsolutions.net _______________________________________________ NumPy-Discussion mailing list -- numpy-discussion@python.org To unsubscribe send an email to numpy-discussion-le...@python.org https://mail.python.org/mailman3/lists/numpy-discussion.python.org/ Member address: arch...@mail-archive.com
